Trained yolo5 on GPR images and got bad results

[Rajat-Mehta]

I have 1200 images (1000 train, 200 test) collected by ground penetrating radar sensor and the goal is to detect different type of defects present in those images. There are 7 classes in total and the defects look like this:

I tried to train your model on my data and it is not working as expected. After 300 epochs I got only 2% mAP and it is not increasing at all. I have made the necessary changes in config files, prepared my data according to your custom data training tutorial. Am not sure what is wrong in my training process.

Do you have any idea or suggestions in order to achieve good mAP on my dataset?

Thanks,
Rajat

[github-actions]

Hello @Rajat-Mehta, thank you for your interest in our work! Please visit our Custom Training Tutorial to get started, and see our Jupyter Notebook , Docker Image, and Google Cloud Quickstart Guide for example environments.

If this is a bug report, please provide screenshots and minimum viable code to reproduce your issue, otherwise we can not help you.

If this is a custom model or data training question, please note that Ultralytics does not provide free personal support. As a leader in vision ML and AI, we do offer professional consulting, from simple expert advice up to delivery of fully customized, end-to-end production solutions for our clients, such as:

• Cloud-based AI systems operating on hundreds of HD video streams in realtime.
• Edge AI integrated into custom iOS and Android apps for realtime 30 FPS video inference.
• Custom data training, hyperparameter evolution, and model exportation to any destination.

For more information please visit https://www.ultralytics.com.

[glenn-jocher]

@Rajat-Mehta I've pushed a commit for better anchor-label introspection before training which may help future training runs.
31f3310

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.